DeepSeek-V3.2 vs Qwen3.6 Plus
Wins 0 of 7 areas
—
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents · Following instructions
Qwen3.6 Plus is the stronger all-rounder.DeepSeek-V3.2 is cheaper.
Scores updated · 39 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software08Qwen3.6 Plus8 of 8 tests
- AgentsCarrying out multi-step tasks on its own03Qwen3.6 Plus3 of 3 tests
- MathCompetition and research-level math03Qwen3.6 Plus3 of 3 tests
- ReasoningHard problems that need careful thinking02Qwen3.6 Plus2 of 3 tests · 1 tie
- Long documentsFinding answers in very long texts02Qwen3.6 Plus2 of 2 tests
- FactsGetting facts right instead of making them up12Qwen3.6 Plus2 of 3 tests
- Following instructionsDoing exactly what it is asked01Qwen3.6 Plus1 of 1 test
Following instructions rests on a single test.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
DeepSeek-V3.2 costs 80% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Qwen3.6 Plus pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+48.1points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+41points ahead
- Short factual questions, answered correctlySimpleQA Verified+16.6points ahead
Where DeepSeek-V3.2 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+6.6points ahead
Every test, side by side
All 39 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingQwen3.6 Plus
- SWE-bench ProQwen3.6 Plus by 4115.656.6+41
- Terminal-Bench 2.1Qwen3.6 Plus by 14.646.861.4+14.6
- LMArena · WebDevQwen3.6 Plus by 99 rating points13621461+99 rating
- Terminal-Bench HardQwen3.6 Plus by 8.335.643.9+8.3
- SWE-bench VerifiedQwen3.6 Plus by 5.773.178.8+5.7
- LiveCodeBench v6Qwen3.6 Plus by 3.883.387.1+3.8
- SWE-bench MultilingualQwen3.6 Plus by 3.670.273.8+3.6
- SciCodeQwen3.6 Plus by 1.838.940.7+1.8
AgentsQwen3.6 Plus
- GDPValQwen3.6 Plus by 14.99.824.7+14.9
- MCP AtlasQwen3.6 Plus by 11.962.274.1+11.9
- τ-Bench V3 · BankingQwen3.6 Plus by 218.820.8+2
ReasoningQwen3.6 Plus
- GPQA DiamondQwen3.6 Plus by 4.28488.2+4.2
- Humanity's Last ExamQwen3.6 Plus by 3.224.627.8+3.2
- CritPttie2.92.9tie
FactsQwen3.6 Plus
- AA-Omniscience · Non-hallucinationQwen3.6 Plus by 48.117.365.4+48.1
- SimpleQA VerifiedQwen3.6 Plus by 16.627.544.1+16.6
- AA-Omniscience · AccuracyDeepSeek-V3.2 by 6.63326.4+6.6
MathQwen3.6 Plus
- IMOAnswerBenchQwen3.6 Plus by 5.578.383.8+5.5
- HMMT Feb. 2026Qwen3.6 Plus by 3.784.187.8+3.7
- AIME 2026Qwen3.6 Plus by 1.194.295.3+1.1
Long documentsQwen3.6 Plus
- AA-LCRQwen3.6 Plus by 573.378.3+5
- LongBench v2Qwen3.6 Plus by 2.259.862+2.2
Following instructionsQwen3.6 Plus
- IFBenchQwen3.6 Plus by 14.560.775.2+14.5
Other results16 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA-OmniscienceQwen3.6 Plus by 23.4-22.50.9+23.4
- Terminal-Bench 2.0Qwen3.6 Plus by 15.246.461.6+15.2
- AA Agentic IndexQwen3.6 Plus by 10.718.329+10.7
- Artificial Analysis Coding IndexQwen3.6 Plus by 10.344.254.5+10.3
- MCPMarkQwen3.6 Plus by 10.23848.2+10.2
- HLE (with tools)Qwen3.6 Plus by 9.840.850.6+9.8
- τ²-Bench Telecom (AA run)Qwen3.6 Plus by 7.190.697.7+7.1
- frontiermath_tier_4_v1Qwen3.6 Plus by 6.22.18.3+6.2
- AA IntelligenceQwen3.6 Plus by 5.521.527+5.5
- HMMT Nov. 2025Qwen3.6 Plus by 4.69094.6+4.6
- Tool-DecathlonQwen3.6 Plus by 4.635.239.8+4.6
- ToolathlonQwen3.6 Plus by 4.635.239.8+4.6
- HMMT 2025Qwen3.6 Plus by 4.490.294.6+4.4
- MMLU-ProQwen3.6 Plus by 3.58588.5+3.5
- τ³-BenchQwen3.6 Plus by 1.569.270.7+1.5
- MMLU-Reduxtie93.794.5tie
Questions people ask
Which is better, DeepSeek-V3.2 or Qwen3.6 Plus?
Qwen3.6 Plus wins all seven areas where both have results: coding, agents, reasoning, facts, math, long documents and following instructions. DeepSeek-V3.2 wins none, but costs 80% less.
Which is better for coding?
Qwen3.6 Plus. It wins 8 of the 8 coding tests both models report; DeepSeek-V3.2 wins none.
Which is cheaper?
DeepSeek-V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens; Qwen3.6 Plus costs $0.50 and $3.00. That makes DeepSeek-V3.2 about 80% cheaper for the same work.
How do you compare the two?
We use the 39 benchmark tests both models have published scores on. The verdict counts the 23 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.